dissertation jiri svoboda - uni-regensburg.de · 2011. 7. 20. · *this chapter was already...

Flavin-Based Photocatalysts

Dissertation

Zur Erlangungdes Doktorgrades der Naturwissenschaft

(Dr. rer. nat.)an der Fakultät Chemie/Pharmazie

der Universität Regensburg

vorgelegt vonJǐŕı Svoboda

aus Prag2007

Promotionsgesuch eingereicht am: 21. August 2007.Promotionskolloquium: 11. September 2007.Die Arbeit wurde angeleitet von: Prof. Dr. Burkhard König.Prüfungsausschuss: Prof. Dr. Jörg Heilmann (Vorsitz),

Prof. Dr. Burkhard König (1. Gutachter),Prof. Ivan Stibor, CSc. (2. Gutachter),Prof. Dr. Ruth Gschwind (3. Prüfer).

ii


Dissertation

Zur Erlangung des Doktorgrades der Naturwissenschaft(Dr. rer. nat.)

an der Fakultät Chemie/Pharmazieder Universität Regensburg

vorgelegt vonJǐŕı Svoboda

aus Prag2007

“Nothing is impossible, if you can imagine it. That’swhat being a scientist is all about.”

Prof. H. J. Farnsworth

Contents

1 Templated Photochemistry: Introduction 11 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Templates Containing a Shield . . . . . . . . . . . . . . . . . 33 Photochemical Reactions in Non-Covalent Assemblies . . . . 124 Complementary DNA Strands as Templates . . . . . . . . . . 225 Chromophore–Recognition Site Templates . . . . . . . . . . . 256 Photochemical Reactions in a Molecular Flask . . . . . . . . . 307 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448 Notes and References . . . . . . . . . . . . . . . . . . . . . . . 44

2 2’-Oxoethyl Flavin Revisited 551 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . 573 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

3 Thiourea-Enhanced Photooxidation 631 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632 Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643 Catalytic Properties . . . . . . . . . . . . . . . . . . . . . . . 724 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Summary 81

4 Experimental Procedures 831 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832 Equilibration Experiments . . . . . . . . . . . . . . . . . . . . 843 Theoretical Computations . . . . . . . . . . . . . . . . . . . . 844 Kinetic Experiments . . . . . . . . . . . . . . . . . . . . . . . 845 Cyclic Voltammetry . . . . . . . . . . . . . . . . . . . . . . . 856 2’-Oxoethyl Flavin (2) . . . . . . . . . . . . . . . . . . . . . . 867 Flavin 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878 2-Nitroaniline 11 . . . . . . . . . . . . . . . . . . . . . . . . . 889 Dinitrobenzene 13 . . . . . . . . . . . . . . . . . . . . . . . . 8910 2-Nitroaniline 14 . . . . . . . . . . . . . . . . . . . . . . . . . 89

vii

11 Compound 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . 9012 Benzyl Carbamate 17 . . . . . . . . . . . . . . . . . . . . . . 9113 Trifluoroacetamide 18 . . . . . . . . . . . . . . . . . . . . . . 9214 Flavin 21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9315 Flavin 22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9516 Flavin 23 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9617 Flavin 26 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9718 Flavin 27 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9819 Flavin 28 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9920 General Procedure 1 . . . . . . . . . . . . . . . . . . . . . . . 10021 Flavin 29 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10022 Flavin 30 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10123 Flavin 31 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10224 Flavin 32 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10325 General Procedure 2 . . . . . . . . . . . . . . . . . . . . . . . 10426 Flavin 33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10427 Flavin 34 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10528 Flavin 35 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10629 Flavin 36 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10730 General Procedure 3 . . . . . . . . . . . . . . . . . . . . . . . 10831 Flavin 37 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10832 Flavin 38 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10933 Flavin 39 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11034 Flavin 40 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11135 Flavin 41 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11236 Flavin 42 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11337 Flavin 43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11438 Flavin 44 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11539 Flavin 45 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11640 Bis-Flavin 46 . . . . . . . . . . . . . . . . . . . . . . . . . . . 11741 Bis-Flavin 47 . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

5 Notes and References 119

A List of Abbreviations 127

viii

Acknowledgement

• From my entire heart, I thank my wife Jaroslava Svobodová for all herlove, support and understanding. Without her, this work and my lifewould never be complete.

• I deeply thank my father, Jiř́ı Svoboda, for his most precious advice,patience and fruitful discussions which helped a lot not only with thenumerous synthetic troubles, but during all hard times in the lab.

• I am very much grateful to “meinem sehr geehrten Laborkollegen”Harald Schmaderer for the pleasant and constructive ambient in thelab, friendly chats, coffee-brewing, and in fact, for all the nice timespent together.

• I thank Giovanni Imperato for his patient help, and also for reliablebelaying, thus avoiding any “Jiri-Pizzas”.

• This work was financially supported by the Graduate College GRK640and by the DFG Priority Programme SPP1118. The support is greatlyacknowledged.

ix

Chapter 1

Templated Photochemistry:Toward Catalysts Enhancingthe Efficiency and Selectivityof Photoreactions inHomogeneous Solutions*

1. Introduction

Photochemical reactions are an important tool in modern synthetic chem-istry. They often lead to products virtually inaccessible by thermal reactionsand proceed along the excited-state pathway. However, it is often difficultto predict and control the outcome of photochemical transformations inhomogenenous solutions where molecules behave rather chaotically. Theirencounters in the thermal bath of Brownian motion occur within statisticaldistribution between all reaction geometries. Nevertheless, a defined mutualgeometry is the origin of regio- or stereoselectivity of a chemical reaction,and improvement of reaction selectivity is therefore an attempt to overcomethe equalising power of thermal fluctuations by additional control elements.Bias in the orientation may come from the reacting partners themselves (e.g.,the Cram and Felkin–Anh rules of 1,2-stereoinduction),1 chiral auxiliaries,2

or selective reagents.3 Orientation of the reactants can be also influencedby the widely used transition-metal catalysts with sophisticated ligands.4

On the other hand, spatial arrangement of molecules and their movementis more or less fixed in the solid state. Reactions in the crystal lattice may

*This Chapter was already published as a review, see: Svoboda, J.; König, B. Chem.Rev. 2006, 106, 5413–5430. For the ease of reading, structures, schemes and referencesare numbered separately from the rest of the Dissertation. Some references are thereforegiven twice, once in Section 8 of this Chapter, and once in Chapter 5 on page 119.

1

2 Chapter 1. Templated Photochemistry: Introduction

therefore ehibit higher selectivity than those in homogeneous solution. Pho-todimerisations of crystalline aromatic or olefinic compounds belong to theoldest known organic photoreactions. In this type of reactions, the crystallattice locks the relative orientation of the photoreactive groups. If the orien-tation is favourable for reaction, reactivity increases, and vice versa. Unlikethe photochemistry in a homogeneous solution, this leads to the selectiv-ity for some of the photoproducts. Schmidt coined the term “topochemi-cal” principle or “topochemistry” for (non)reactivity determined by a lim-iting distance between the reactive groups.5–7 Although this model foundwidespread acceptance, many exceptions violating this concept were knownfrom the very beginning.8 Later, AFM techniques enabled experimental elu-cidation of solid-state photochemistry. This showed that the supramoleculararrangement of molecules in the crystal plays a more important role for reac-tion control than simple alignment of double bonds. Long-range molecularmovements within crystal upon photochemical reaction and even, althoughrare, topotactic single-crystal to single-crystal reactions were found. Thework is comprehensively covered by recent reviews.9 Reactions of inclusioncomplexes are a variation of the solid-state photochemistry topic.10 Here,co-crystals of a host compound and the starting materials of a photochemicalreaction are used. Supramolecular arrangement may control the regio- andstereoselectivity of the photoprocess. Enantionselective photochemical con-version of chiral crystals into optically active products has been described.11

Different approaches utilise zeolites as supramolecular hosts for photoreac-tions.12 Internal complexation, or intra-crystalline adsorption, occurs bydiffusion of the guest into the channels and cavities of the zeolite crystaland is size and shape selective. Complexation of organic compounds mayreversibly depend on temperature. The geometry of zeolite restricts confor-mation and orientation of included guests and their reaction partners, lead-ing to more selective reactions. In the absence of any low-energy electronicstates of the zeolite, photoreaction occurs only with the included guest.

The common disadvantage of solid-state photoreactions is the difficultprediction and control of reaction selectivity. Finding the suitable crys-tal, co-crystal, or inclusion complex for the desired regio- or stereoselectiveoutcome of a given reactions remains a challenge. Therefore, an attractivestrategy is to transfer the topochemical control from the solid to homoge-neous solution using suitable templates. Such reactions are easier to analyse,design and optimise. We wil discuss recent attempts in this field in the fol-lowing sections.

The scope of this review is limited to templated reactions in a homo-geneous solution which are initiated or mediated by light. By template wemean a host molecule (sometimes containing a sensitiser) of various con-struction (a rational design, cyclodextrin, cucurbituril, DNA, etc.) whichreversibly binds a substrate and typically does not become part of the prod-uct. The substrate undergoes a truly photochemical process or at least a

2. Templates Containing a Shield 3

process which is light-mediated. Reactions where light is only required topre-form a species which undergoes a thermal process are therefore not in-cluded. In the ideal scenario the template should act as a catalyst and bepresent in subequimolar amounts, but this aim could not be always achieveddue to (i) low association constants, (ii) undesired photochemical reactionsin the unbound state which often competed with the desired reaction, or(iii) inhibition of the template by the reaction product. To illustrate theprogress and development in the rapidly advancing field, we extend the dis-cussion of catalytic processes by closely related non-catalytic examples andauxiliary approaches where the template is covalently bound to the substratebut could be easily regenerated. We focus on reactions in true homogeneoussolutions only and do not consider any confined media, such as crystals,co-crystals, any kinds of solids, polymers, solid phase + gas reactions, reac-tions in zeolites, micelles, membranes, compartmented solutions, aggregates,or reaction in and of liquid crystals. We refer the interested reader to reviewsand examples discussing the topics exceeding the scope of this review13 aswell as to recent reviews related to the discussed topic.14

The division of the sections is somewhat artificial as they partly overlapand some of the work could be featured in several of them. However, wetried to divide them in a way to make reading and understanding easier.The survey begins with photoreactions in which selectivity was achievedby shielding one face of a reactant. Not all the discussed examples useda catalytic template, but as we reasoned above, they illustrate the concep-tual development over the last years. The next group of examples mimicstopochemical reactions in the solid state by non-covalent assemblies wherethe substrates are pre-organised in a defined way by the host molecule. Ex-amples using DNA as a template are followed by contributions featuringcovalently bound sensitiser and a substrate-binding site to enhance the ef-ficiency of energy- and electron-transfer processes. Finally, photoreactionsin molecular flasks or soluble cavities establish a relationship to reactions inzeolites or inclusion complexes.

2. Templates Containing a Shield

This section describes those examples which bind a substrate by non-cova-lent interactions to distinguish between different sides of attack and thusdirect the course of a photochemical transformation by the presence of asteric hindrance.15,16 Some of the reactions are bimolecular and some in-tramolecular. However, in all cases, direction of the approach is not governedby the size and shape of the substrate itself but by stereodifferentiation ofits two faces by the template’s shield (stereoexclusive approach).

Pioneering work in this area was carried out by Mori et al. who studieda reaction between compound 1, which contains a covalently bound imide


binding site and a coumaring group, and 3-butylthymine 2 (Scheme 1).17,18

N

NH

OO

Bu

OONH

O

O

O

O

N

NH

OO

Bu

OONH

O

O

O

O

.1 2 3

hν

benzene

40%

Scheme 1. Stereoselective cycloaddition of coumarin and thymine

We are aware of the fact that in this example compound 1 did not actas a catalyst because after the photochemical reaction, cycloadduct 3 wascovalently bound to the template which could not be reused. However,we wish to include this example for “historical” reasons to illustrate theprogress in the field which is dynamic and quickly developing. The presenceof the covalently bound recognition site increased the reaction rate and thecis-syn/cis-anti ratio up to 96:4

A similar—still noncatalytic—example from Bach et al. described thestereoselective Paternó–Büchi reaction of 3,4-dihydro-1H -pyridine-2-one 4with a covalently host-bound benzaldehyde 5.19,20 Substrate 4 was boundto chiral host 5 by two directional hydrogen bonds, and its enantiotopicfaces were thus differentiated (Scheme 2). Diastereoselectivity depended onsolvent polarity in accordance with the expectation: polar solvents such asacetonitrile disturbed formation of the complex and lowered the diastere-omeric ratio, while non-polar solvents such as benzene favoured formationof the complex. Diastereomeric ratios up to 95:5 and up to 56% yield wereobserved (toluene, −10 ◦C). The importance of hydrogen bonding for thesteroselectivity was proved by a control experiment with the N -methylatedhost 6, which does not allow face discrimination and showed no stereoselec-tivity.

The next steps toward efficient catalytic asymmetric induction were chi-ral templates containing benzoxazole (host 7) or menthol (host 8) as stericshields (Scheme 3). Templates 7 and 8 were used in an intramolecular [2+2]photocycloaddition of substituted 2-quinolones 9 (Scheme 4).21 The men-thol residue did not sufficiently discriminate the two sides of the substrateand the ee values remained low when using the template 8. The benzoxazole-containing host 7 proved superior, and high yields and ee values could be


N OO

O

H

OR

OONO

HH

ON

O

NH

OOO

O

NH

O

4 . 5R = HMe

56

hν

solvent

Scheme 2. Stereoselective Paternó–Büchi reaction and its predominant product

NH

O NO NH

O OO

7 8

Scheme 3. Chiral templates used for catalytic asymmetric induction

N O

O

NO

HH

NO NH

O

O H

H

n = 13

. 7

. 7

NH

O

HO

H

hν

toluene

( )n

9a9b

10a 10b

up to 77%up to 93% ee

up to 88%up to 88% ee

Scheme 4. Intramolecular [2+2] cycloaddition


achieved by use of more than one equivalent of the host, thus forcing associ-ation with the substrate. The four-membered ring in 10a was formed withup to 79% yield and 84% ee (77% yield and 93% ee of ent-10a is reportedfor the reaction at −60 ◦C with 2.6 eq. of host ent-7) and the six-memberedring in 10b with up to 88% yield and an 88% ee (−15 ◦C, toluene, 1.2 eq.of host 7).

Photocycloaddition of 4-methoxy-2-quinolone 11 with various olefins 12was studied in detail.22 Cycloaddition of 11 with symmetric olefin 12a inthe presence of chiral host 7 yielded a single product 13, which was formedwith 61% yield and 92% ee (Scheme 5). Non-symmetric olefins 12b–f gave

NH

OMe

O NHO

H

MeOR2 R

1

NH

O

H

MeOR1 R

2

R2

R1

12a12b12c12d12e12f

EtHHHHH

R2 =

hν, -60 °Ctoluene

11 12

R1 = EtCH2CH2CH2OHCH2OAcOAcPhCOOMe

+ enantiomer + enantiomertemplate 7 13a 13b

Scheme 5. [2+2] Photocycloaddition of 4-methoxy-2-quinolones; See text for yields andee values

a mixture of two diastereomers 13. Olefins 12b–c,f showed diastereoselec-tivity in favour of isomer 13a (ratios 13a:13b were 90:10 or higher, yields80–84%) with ee values from 81% to 92%. Styrene 12e favoured forma-tion of the corresponding endo-isomer 13b (ratio 13a:13b was < 5:95, yield29%) with 83% ee. Olefin 12d yielded a 63:27 mixture of products 13a and13b (overall yield 89%). However, both of the products were formed highlyenantioselectively (13a with 93% ee, 13b with 98% ee). Replacing the chiralhost by its enantiomer resulted in product enantioselectivity reversal.

Enantioselectivity of [2+2]-photocycloaddition of protected 4-(2’-amino-ethyl)-quinolones 14 with acrylates 15 could also be successfully manipu-lated by the template 7 (Scheme 6, top).23 In its presence, the exo products16 were isolated with 44–86% yield and 70–81% ee. Higher ee values wereobserved at lower temperatures. Superstoichiometric amounts of the tem-plate 7 (2.5 eq. based on the quinolone 14) had to be employed in order toovercome the self-association of 14 to which the incomplete chirality trans-fer was attributed. Variation of quinolone 14 or acrylate 15 concentrations


NH

O

NBocR1

R2

O

O

O

O

NH

O

NBoc

R1 COOMeR2

H

O

NH

O

NBoc

R1O

H

R2 = HMeEt

15a15b15c

R1 = CH2Phtert-Bu

14a14b

NH

O

N

OR

Ph

NH

O

H

N RO

Ph

+

15d

16a-c

16d

host 7toluene-60 °Chν

44 to 86%, 70 to 81% ee

host 7toluene-60 °Chν

17a17b17c

R = HMeEt

18a18b18c

46%, 92% ee53%, 76% ee55%, 74% ee

Scheme 6. Enantioselective inter- and intramolecular cycloaddition of 4-(2’-aminoethyl)-quinolones; See text for individual yields of the intramolecular cycloaddition


had no influence on the observed ee value. At −60 ◦C, there were no de-tectable side reactions due to hydrogen abstraction. Compounds 17 cyclisedin the presence of the chiral host 7 to render the tetracyclic products 18 inan intramolecular fashion (Scheme 6, bottom). Yields of the reaction werelowered by competing intramolecuar oligomerisation. However, observed eevalues were higher than the aforementioned intermolecular reaction becausethe oligomerisation reduced the extent of racemic intramolecular reactionsnot templated by the host. This is best illustrated by the reaction of 17awhich gave product 18a with the lowest yield (46%) but highest ee value(92%).

The template 7 was successfully used for the stereoselective Diels–Alderreaction of photochemically generated (E )-o-quinodimethane with alkenes24

and for enantioselective radical cyclisation reaction of 4-(4’-iodobutyl)quin-olones.25

The chiral host 7 found another application for the stereoselective [4+4]photocycloaddition of 2-pyridone 19a to cyclopentadiene 20 (Scheme 7).26

Upon irradiation in the presence of the chiral host ent-7, a mixture of di-

NH

ONH

O

NH

O

-50 °Ctoluene

template ent-7

35%87% ee

+

41%84% ee

19a

20

21a 21b

hν

Scheme 7. [4+4] Photocycloaddition of 2-pyridone to cyclopentadiene

astereomeric products 21, which were formed with significant ee values, wasobtained.

Intramolecular [4π] cyclisation of 2-pyridone 19a was considered, too.26

Since the reaction is known to proceed with low yields, 4-substituted pyri-dones 19b,c, which undergo cyclisation more readily, were investigated. Thereaction rate and yield increased with temperature, but on the other hand,low temperature was required for high association of the substrate with thehost. The highest stereoselectivity (23% ee, 51% yield) was observed for4-benzyloxy-2-pyridone 19c at −20 ◦C (Scheme 8).

Another stereoselective photochemical reaction described was the Nor-rish–Yang cyclisation.27 Upon irradiation in the presence of chiral host 7 or8, imidazolidinones 22 yielded a mixture of stereoisomers 23 (Scheme 9).The exo isomer 23a prevailed in the reaction mixture with a ratio of about4:1 without a significant influence of the reaction temperature. The ee val-ues of the exo isomer 23a depended on temperature and increased at lower


NH

O

R

NH

OH

H

R

R = HOMeOCH2Ph

hν

toluenetemplate ent-7

+

19a19b19c

enantiomer

Scheme 8. [4π] Cyclisation of 2-pyridones

NNH

O

PhO

NH

O

N

HOHPh

NH

O

N

HOH Ph

toluenechiral host

+

+ enantiomers22 23a 23b

hν

host 8,host 7,host 7,

30 °C:-45 °C:-45 °C:

86%,77%,60%,

5% ee,26% ee,60% ee,

exo:endo 88:12exo:endo 79:21exo:endo 80:20

Scheme 9. Norrish–Yang cyclisation

temperatures. Similar to the aforementioned example, the benzoxazole-containing host 7 induced higher ee values than its menthol analogue 8 be-cause of better face differentiation. The best results reached 60% ee (toluene,−45 ◦C). Within the error limit, the enantiomeric chiral host ent-7 deliv-ered the same ee, predominantly giving the product ent-23a. The ee valuesof the minor endo diastereomer 23b were not always determined, but theywere usually lower than for the exo isomer 23a.

The chiral template ent-7 was also used for stereoselective [6π] photocy-clisation of compound 24a.28 The cyclohexene ring located in the vicinityof the shield preferred turning away from it during the photochemically al-lowed conrotatory ring closure due to steric reasons (Scheme 10). Therefore,the zwitterionic intermediate was formed stereoselectively in the presence ofthe template. Upon protonation by the host, which acted as a Brønstedacid, the zwitterionic intermediate tautomerised to yield the products 25a.Since the configuration at carbon atom C10a was already fixed, protonationof carbon atom C6a could give rise to the cis product cis-25a or to the transproduct trans-25a. Dissociation of the intermediate from the template,which was required for the protonation, opened its re face (with respect tothe carbon atom C6a), and preference for the trans product was therefore


O NH

A

O NH

H

HO N

H

H

H

N NO

O

HO N

H

+- H+

A = CN

24a24b

O NH

HH

- +

+

24

ent-7

cis-25a trans-25a

10a

6a

up to 45% ee up to 57% ee

ent-7 . 24a

hν

Scheme 10. Enantioselective [6π] photocyclisation; Yields and ee values depend on con-ditions of the reaction, see text for discussion

observed. Thus, the major product trans-25a (trans:cis ratio up to 73:27 intoluene at −55 ◦C) was obtained with up to 57% ee (toluene, −55 ◦C). Theminor cis product cis-25a was obtained with up to 45% ee (toluene, 35 ◦C).Similar experiments with the analogous 24b rendered very low ee values,and its cyclisation was therefore not studied in detail.

Prochiral quinolone 26 cyclised to a chiral pyrrolizidine 27 by a pho-toinduced electron-transfer reaction (PET) in the presence of chiral host28 (Schemes 11 and 12).29 The chiral host 28 fulfilled three conditions:(i) it contained an efficient sensitiser of the electron-transfer process in closeproximity to the substrate, (ii) it bound the substrate by two directionalhydrogen bonds, and thus (iii) it differentiated its two enantiotopic faces.The reaction proceeded successfully in the presence of 5–30 mol % of thechiral host 28, reached ee values up to 70%, and gave yields up to 64%(higher values were observed with higher amounts of template). Again, useof the enantiomeric form of the host led to reversion of the enantiomericpreference. The results were also compared to a reaction where a mixtureof the benzoxazole chiral host 7 and p-methoxybenzophenone was used, andthe importance of sensitiser proximity was thus proved.

The intramolecular enone–olefin [2+2] photocycloaddition of quinolone


NH

O

N

NH

O

N

NH

O

N

NH

O

N

NH

O

N

NH

O

N

NH

O

N

NH

O

N

1

2

+

- 1e- - H+

+ 1e-

toluene-60 °C

hνchiral hostPET

26 27 up to 70% ee, 64% yield

+

Scheme 11. Enantioselective photochemical formation of pyrrolizidine 27 and its pro-posed mechanism; Key: (1) photoinduced electron transfer, (2) back electron transfer fromtemplate

NH

NO

O

ONH

NO

OH

O

NO

N

H

26. 28+ enantiomer

28

Scheme 12. Chiral host 28 used for enantioselective cyclisation and its function


29 reported by Cauble et al.30 showed that in the absence of a rigid aro-matic shield, some enantioselectivity is still achieved in catalytic templatedphotochemistry. The “sensitising receptor” 30 contains a complementaryhydrogen-bonding pattern to bind the substrate 29 (Scheme 13). The pres-

N

NH

O

O

NH

O

CH3

O

NH

O

N

N

O

O

N

O

R

CH3H HO

29 . 3030

* *

Scheme 13. “Sensitising” receptor 30 used for stereoselective [2+2] photocycloadditionof quinolone 29 and its complex with the substrate; R = 4-C6H4(C=O)Ph

ence of the receptor 30 increased the efficiency of the cyclisation reaction andallowed for its stereoselective course at lower temperatures upon which theasymmetric induction strongly depended (highest ee values of ca. 20% wereachieved at −70 ◦C). The reaction required only catalytic amounts of the re-ceptor 30. Experiments with 1:1 and 4:1 substrate 29:receptor 30 mixturesgave comparable enantiomeric excess, which indicates that the observed levelof asymmetric induction indeed results from the intrinsic enantiofacial biasconferred by association of quinolone 29 to the sensitising receptor.

3. Photochemical Reactions in Non-Covalent As-semblies

As already stated in the Introduction (page 1), the spatial arrangement ofmolecules in the crystal is defined by the rigid crystal lattice and selectivephotochemical reactions could be thus achieved in the solid state. We con-tinue our review with examples where supramolecular templates mimic thespatial fixation in crystals by pre-organisation of the reaction componentswithin a non-covalent assembly in an advantageous and better defined ge-ometry, thus leading to an increase in reaction rate and selectivity.

In the first part, development of a covalent substitute for the dimerisationof cinnamic acid is described. In this case, the two components were pre-

3. Photochemical Reactions in Non-Covalent Assemblies 13

organised within a covalent assembly which could be, however, consideredtemporary and reusable, and we therefore include this work as well. Thesecond part discusses truly non-covalent assemblies and begins with systemswhich employ the template effect of alkali or alkaline earth metal cations oncoronand systems, and concludes with systems which utilise recognition ofdi- or triaminotriazine by barbiturate and other suitable molecules.

3.1 Temporary Covalent Pre-organisation

In the [2.2]paracyclophane system 31a, the two reactive parts of the moleculewere kept in close proximity (ca. 3 Å), thus resembling the arrangementin the crystal.31–33 It was shown that the [2π+2π] cycloaddition of thisand vinylogous cyclophane systems 31a–c to cyclobutane derivatives 32proceeded with excellent yield and stereoselectivity (Scheme 14).

n

n

COOEt

COOEtn

COOEt

COOEt

100%> 90%83%

32a32b32c

n = 123

31a31b31c

-90 °Csolvent

hν

Scheme 14. Photochemical synthesis of [n]ladderanes

Interestingly, the effect of the pseudolattice extended beyond its imme-diate vicinity.34 However, this advantage could not be employed in catalyticreactions nor could the template be regenerated, as the reacting centerswere connected by C–C bonds. Therefore, a similar system 33 in whichthe template served as an auxiliary and could be recycled was prepared.35

The rigid cyclophane lattice bore two connection points to which the actualreaction substrates 34 bound. Once bound in a conjugate 35a, they couldundergo the desired photochemical reaction yielding 35b. After removal ofthe β-truxinic product 36, which was formed exclusively, the rigid cyclo-phane space 33 was regenerated. The amide bonds were easy to form andeasy to cleave and thus served as a temporary linkage between the template33 and the substrate 34. All individual steps of the cycle ran with good toexcellent yields (Scheme 15).

3.2 Non-Covalent Intramolecular Pre-Organisation

The efficiency of dimerisation of anthracene, incorporated in a macrocycliccrown ether or tethered by an ethylene glycol chain, was influenced by com-


NH2

NH2

Cl

ONH

NH

O

O

NH

NH

O

O

NH2

NH2. 2 HClHOOC

HOOC

+ 2

80%

76%98%

97%

+

33 34 35a

36

i)

ii)

iii)

iv)hν

35b

Scheme 15. Temporary use of the cyclophane unit for topochemical reaction control insolution; Conditions: (i) 1,4-dioxane, r.t., 24 h, (ii) hν, acetone, 7 h, (iii) conc. hydrochloricacid, reflux, 24 h, (iv) solid potassium hydroxide

plexation of guest ions.36–41 In the simpler cases, efficiency and regioselec-tivity of the cycloaddition responded to only one chemical input: either tothe presence of alkali metal cations38 or to the presence of alkaline earthmetal cations.39

The regioselectivity of the photocycloaddition of two anthracene unitsin 37 was subject to the presence of sodium ions.38 If present, a symmetricproduct 38 (Scheme 16) was obtained (quantum yield ΦR = 8.3 × 10−3in methanol), while in the absence, a non-symmetric photocycloadditionproduct 39 was obtained with much lower quantum yield (ΦR = 1.9× 10−4in methanol). Compound 38 is not thermally stable, slowly reverting to thestarting material at room temperature. Potassium ions could also be usedto yield the symmetric regioisomer 38, but the quantum yield of the process(ΦR = 1.8 × 10−3 in methanol) is lower.

The regioselectivity of photocycloaddition of bis-anthracene 40 respond-ed to the presence of alkaline earth metal ions.39 Upon irradiation, threeregioisomers were formed; isomer 41d was not detected (Scheme 17).42,43

With increasing ion radius of the alkaline earth metal ion present, preferencefor regioisomer 41c was observed at the cost of isomer 41a (the ratio of 41cincreased from 17% with magnesium ions to 46% with barium ions, the ratioof 41a decreased from 81% with magnesium ions to 54% with barium ions).Formation of isomer 41b was prohibited in all cases (relative isomer ratio 2%with magnesium ions down to 0.1% in the presence of barium ions). Alkalimetal ions did not influence the ratio of the isomers. Identical experiments


O

OO

O

O

O

OO

O

O

O

O

O

O O

O

O

OO

O

O OO

O O

OOO

OONa+ Na+

O

OO

O

O

O

OO

O

O

Na+

Na+

hν366 nmmethanol

NaClO4methanol

37 39

stable, yield 80%

hν366 nmmethanol

3837 . 2 Na+

Δ

not stable

Scheme 16. Cation-templated dimerisation of anthracene


NH

O OO

O NH

O

O

O

O NH

NH

O

O=

hνacetonitrile

40

41a 41b 41c 41dnot observed

M+

Scheme 17. Dimerisation of two glycol-linked anthracenes; All possible isomers shown,product ratio depended on metal cation present (see text). Preparative yields were notpublished

were carried out with anthracenes linked in the 2-position. However, inthis case, no effect of alkali or alkaline earth metal ions on the regioisomerdistribution was observed.

It was also possible to construct a system which exhibited the logical“or” operation and responded to two different stimuli.41 Molecule 42 con-tained two anthracene units which dimerised upon irradiation, two glycolchains for binding of alkali metal ions, and a 2,2’-bipyridyl unit for bindingof transition-metal ions (Scheme 18). Upon irradiation of compound 42 inthe presence of sodium or mercury(ii) ions, the quantum yield of dimeri-sation increased 2-fold as opposed to the reaction in the absency of anycation template (ΦR = 0.23 in the absence of template, 0.37 in the pres-ence of mercury(ii) ions, and 0.43 in the presence of sodium ions or bothof the cations, measured for compound 42a in acetonitrile, and ΦR = 0.19


OOON

O O O Nn n

N

N

O

O OO

OOn

n

43a,b

n = 01

42a42b

hνΔ

Scheme 18. Dimerisation of anthracene dependent on two different stimuli; The productis not thermally stable, and the efficiency of the photochemical reaction in the presenceof metal cations was characterised by quantum yields (see text)

in the absence of templates, 0.26 in the presence of mercury(ii) ions, 0.29in the presence of sodium ions, and 0.32 in the presence of both cations,measured for compound 42b in acetonitrile). The quantum yield remainedhigh when both stimuli were present, as required for an “or” gate. The rateof thermal dissociation decreased and the stability of the photoproduct 43increased in the presence of sodium ions. The dissociation constant droppedfrom 97 × 10−6 to 29 × 10−6 s−1 for compound 43a and from 202 × 10−6to 2.9 × 10−6 s−1 for compound 43b. On the other hand, the presence ofmercury(ii) ions did not significantly influence the product stability.

3.3 Non-Covalent Intermolecular Pre-Organisation

A diaminotriazine–barbiturate-based assembly was successfully used for di-merisation of cinnamates and stilbenes.44,45 The hydrogen-bonding pat-tern of the diaminotriazine present in molecules 44 was complementaryto that of barbituric acid 45. In its presence, several hydrogen bondeddimers and trimers can form. However, only one of them (complex 46)locked the two cinnamates or stilbenes 44 in a suitable geometry for pho-tocycloaddition (Scheme 19). The presence of the barbiturate template 45increased the reactivity and induced selectivity for some of the product iso-mers (Scheme 20). For dimerisation of cinnamates 44a and 44b, formationof β-truxinate products 47a,b was preferred (quantum yield increased from


N N

NO

NH

NH

R1

R2

R3

NH NH

O

O O

R1

R1

R1

R1 R1

R1R1

R1

R1 R1

= =

44a44b44c

45R1 = R2 = R3 =COOMeCOOMePh

Mei-Pri-Pr

H4-t-Bu-PhH

+K1

+

K2 + +

46

Scheme 19. Dimerisation of cinnamate and stilbene in a non-covalent assembly


R1R1

ArAr

NHN

N N

NHR3

OR2

47a47b47c

R1 = R2 = R3 =COOMeCOOMePh

Mei-Pri-Pr

H4-t-Bu-PhH

ArR1

R1Ar

R1R1

ArAr

ArR1

R1Ar

NHN

N N

NHR3

OR2

Ar = R1 = R2 = R3 =Ph i-Pr H

Ar =

47d

47e

47f

Scheme 20. Products of cinnamates and stilbene 44 photodimerisation; See text forquantum yields of their formation

0.7×10−3 to 2.3×10−3 for cinnamate 44a, and from 0.1×10−3 to 0.8×10−3for cinnamate 44b). In the case of the stilbene 44c dimerisation, forma-tion of all products (47c–e) apart from 47f was enhanced (quantum yieldincreased from 0.8 × 10−3 to 4.6 × 10−3 for the products 47c–e altogetherbut decreased from 0.7 × 10−3 to 0.5 × 10−3 for product 47f).

A melamine–barbiturate assembly (Scheme 21) was successfully usedfor dimerisation of fullerenes.46,47 Two fullerene–barbiturate conjugates 48were bound by the melamine host 49 and efficiently underwent dimerisationupon irradiation. In the absence of the melamine host 49, no dimerisationwas observed.

The effect of diaminotriazine–barbiturate and glycol–metal cation bind-ing was combined in one molecule 50 (Scheme 22) to further enhance theefficiency of cinnamate dimerisation.48 Accordingly, the cinnamate was“equipped” with both a diaminotriazine substituent and a polyethylene gly-


O

NH NH

O O

NH NH

O

O O

C60 C60

N NH H

O

O

N N

NNC8H17

H

NH H

O

NH

C8H17

O

N NH

O O

H

482 49.

hν

C60 C60

Scheme 21. Dimerisation of fullerenes in a melamine–barbiturate assembly; Yield of thereaction was not given

col chain. Quantum yields of dimerisation and product selectivity werestudied in the presence or absence of barbiturate 45 and in the presenceof various metal cations. For example, efficiency of dimerisation (taken asyield of dimers after a certain time of irradiation) increased twice in thepresence of 0.5 eq. of barbiturate 45 or potassium ions. When both of thetemplates were present, a 5-fold increase in dimerisation efficiency was ob-served, which could be explained by co-operative effect of the two templates.The most significant effect was observed in the presence of barium ions. In-terestingly, the presence of barium ions during irradiation hardly influencedthe selectivity for product 51a and 51b, although the quantum efficiencyincreased ca. 300 times (from 3.6 × 10−5 to 9.1 × 10−3 for product 51a andfrom 3.7 × 10−5 to 1.0 × 10−2 for product 51b). On the other hand, thepresence of the barbiturate template 45 increased quantum yield only ca.twice (from 3.6 × 10−5 to 6.4 × 10−5 for product 51a and from 3.7 × 10−5to 5.3 × 10−5 for product 51b) but increased the product ratio 51b:51afrom 1:1 to 5:1. This indicated that the main effect of the non-directive ionchelation by polyethylene glycol (as opposed to hydrogen bonding betweendiaminotriazine and barbiturate) was to increase the local concentration ofthe reactants. The flexibility of the glycol chain probably prevented trans-fer of geometrical information from the metal ion coordination site to thephotoactive moieties.

Photodimerisation of coumarins 52 was successfully influenced by a sym-metric ditopic receptor 53 containing two identical hydrogen-binding recog-nition sites (Scheme 23).49 Coumarin 52 forms a 2:1 complex with template53. Unlike free coumarin 52, which yielded a 60:40 mixture of syn andanti coumarin photodimer, irradiation of complex 522·53 led exclusively toformation of trans-syn head-to-head dimer 54. The preference for the trans


O

O O

O O

O

N N

NO

NH2

NH

=

hν

=

Ba2+

50 45

45

51a 51b

+ + ...

502 45 Ba2+. .

Scheme 22. Dimerisation of cinnamates influenced by two supramolecular interactions;See text for quantum yield of the products


isomer was explained by steric effects. Interestingly, the quantum yield ofcoumarin 52 dimerisation was lower in the presence of the template 53(Φ = 0.03, ca. one-half of the quantum yield observed for free coumarin 52)because of template shielding. However, this shielding effect increased thestability of the dimer, protecting it from the backward reaction. The im-portance of hydrogen bonding for photodimerisation efficiency was provedby template analogues with missing or methylated recognition units.

4. Complementary DNA Strands as Templates

In this section we summarise photochemical reactions (ligations of deoxy-nucleotides) whose reactants required pre-organisation by the complemen-tary DNA strand.

Vinyldeoxyuridine (VU) base was incorporated into the oligodeoxynucle-otide strand 55 by automated synthesis, and was connected to another por-tion of the strand (56) by a [2+2] cycloaddition.50 However, in the absenceof the template, represented by the complementary DNA strand 57, no re-action occurred. In the presence of the complementary oligodeoxynucleotide57, both components 55 and 56 bound to the template and underwent cleanand efficient reaction upon irradiation at 366 nm to yield the ligated prod-uct 58 (Scheme 24). The ligation between the 5’-terminal VU residue and3’-terminal C residue is selective. There is no reaction the the 5’-terminalVU and 3’-terminal A or G residue. The process was found to be reversibleand the back reaction proceeded rapidly at 302 nm. Products of the cleav-age reaction could be irradiated again, and the cyclobutane product was

H NN

HN

O

N HN

HN O

O

OC6H13

HO

O

OC6H13

H

O O

OC6H13

Oc6H13

O OH H

522 53.

hν

trans-syn head-to-head

54

acetone

100%

Scheme 23. Ditopic template-assisted photodimerisation of coumarin

4. Complementary DNA Strands as Templates 23

N

N

O

NH2

N

NH

O

O

TGTGCC VUGCGTG

ACACGG ACGCAC~

N

N

O

NH2

N

NH

O

O

3' 5'

5' 3'

TGTGCC VUGCGTG

ACACGG ACGCAC~

58 . 57

hν

hν'

3' 5'

5' 3'

5556

57 up to 96%

Scheme 24. Oligodeoxynucleotide-directed photoligation

re-formed. Efficiency of the reaction was proven by the simultaneous liga-tion of five oligodeoxynucleotides 59 (4× VUGTGCC, 59b; 1× CGCAGC,59a) bound to one template 60 (Scheme 25). Again, the expected ligationproduct 61 was cleanly formed.

61 60.

CGCAGC

GCGTCG

UGTGCC

ACACGG

UGTGCC

ACACGG

UGTGCC

ACACGG

UGTGCC

ACACGG

302 nm366 nm

CGCAGC

GCGTCG

VUGTGCC

ACACGG

VUGTGCC

ACACGG

VUGTGCC

ACACGG

VUGTGCC

ACACGG

~ ~ ~ ~

~ ~ ~ ~(59a + 59b4) 60.

Scheme 25. Simultaneous ligation of five vinyldeoxyuridine-containing oligodeoxynu-cleotides; Conversion and reversion are quantitative

Later, the idea was extended to the synthesis of branched oligonucle-otides.51,52 A new, more reactive nucleotide—5-carboxyvinyldeoxyuridine(CVU)—was employed (Scheme 26) which enabled shorter irradiation timeduring the ligation reaction. Similarly to the previous example, the CVU-containing oligoeoxynucleotide 62 was efficiently ligated to a C residue mod-ified by a polyA chain 63 to yield product 64 (Scheme 27) while bound tothe complementary oligodeoxynucleotide strand template 65.

The CVU residue was also used for efficient photoligation to a 3’-terminal


N

NH

O

ON

N

NH2

O

H

H H

COOH

3' 5'

5' 3'N

NH

O

OHOOC

N

N

NH2

O hν

3' 5'

5' 3'

Scheme 26. Structure of the 5’-terminal CVU residue and the photocycloaddition productwith a 3’-terminal C residue

ACACG

5'

3' 5'

3'

CVUGCGTG

ACGCAC

3'hν

ACACG

5'

3' 5'

3'

UGCGTG

ACGCAC

3'TGTG

AAAAAACTGTG

AAAAAAC

~ ~

65(63+62) . 64 . 65

1 h

97%

Scheme 27. Synthesis of a branched oligodeoxynucleotide strand

C or T residue in the presence of the complementary RNA strand, thusforming a hybrid duplex.53 Sensitivity of the photoligation efficiency tomismatches in the sequence of the complementary RNA strands was thensuccessfully used for detection of RNA single-point mutations.

The DNA photoligation methods were extended by the use of α-5-(cyano-vinyl)deoxyuridine at the 3’ terminus,54 contrary to previous examples wherethe modified base was always located at the 5’ terminus. It was predictedby molecular modelling methods that the α anomer would interact betterwith the 5’-terminal T residue better than the β anomer. Again, irradiationat 366 nm quickly and efficiently yielded the ligated oligodeoxynucleotidesin the presence of the complementary DNA strand. The photoligation ef-ficiency depended on sequence specificity; in the case of single mismatchedoligodeoxynucleotides, conversion after identical time periods reached only9–15%. The method was verified by synthesis of a branched nucleotide withT residue at the ligation point modified by a polyA chain, similar to theprevious example.

It was also possible to prepare a photoreactive nucleotide based on adeno-sine instead of uridine.55 In this case, 7-carboxyvinyl-modified 7-deaza-2’-deoxyadenosine 66 (Scheme 28) located at the 5’ terminus reacted with a

5. Chromophore–Recognition Site Templates 25

3’-terminal T or C residue in the presence of a complementary DNA strand.Photoligation was extremely rapid (93% yield within 5 min) and selective

O N

N

N

NH2

OH

OH

HOOC

66

Scheme 28. Structure of 7-carboxyvinyl-7-deaza-2’-deoxyadenosine

for the 3’-terminal C or T residues. A or G residues did not undergo reac-tion. The photoreversibility of the process was similar to the aforementionedexample.

Another option is connecting two DNA strands by photocycloadditionof anthracenes.56 Anthracene units were linked to the 5’ or 3’ termini ofthe strands by a hexamethylene or trimethylene chain (Scheme 29, com-pounds 67). The sequences were designed to stack the anthracene units inthe presence of the complementary DNA strand 68a. Upon irradiation, theasembled anthracene units undergo photodimerisation and connect the twostrands. The relative yields of the anthracene photodimer after 30 min were67d+67a (1.46) > 67b + 67a (1.00) > 67b + 67c (0.33) > 67d + 67c(0.24).57 Photoligation efficiency of 67b with 67a was studied in the pres-ence of several longer (68b–d) or mutated (68e–g) templates. Although aone-residue-long gap (template 68b) was tolerated (relative yield 1.1), theligation efficiency rapidly decreased with increasing length of the gap (rela-tive yield 0.16 with template 68c and 0.06 with template 68d). The processis sensitive to one-base displacement in a position-dependent manner. Effi-ciency decreased in the order 68a (1.00) > 68e (0.35) > 68f (0.19) > 68g(0.00), which shows that a base pair mismatch in the binding site region(68g) reduces the efficiency more than mismatches between the two bindingsites. No ligation was observed with a half-scrambled sequence template.The order of ligation efficiency co-incided with the duplex thermal stabilityorder.

5. Templates with a Covalently Bound Chromo-phore and Recognition Site

The efficiency of a photochemical reaction can be often enhanced by a suit-able sensitiser as an additive. The idea to covalently attach the sensitiser toa substrate-binding site was sucessfully used for efficient oxidation of ben-


OP

O

O O

NH

O

3 ArCGCTAA5' 3'

NH

OP

O

OH

O O

4 Ar

O

OP

O

OO

NH

O

6Ar

OP

O

OO

NH

O

3Ar

TTAGGGTTAGCG

TTAGGGTTTAGCG

TTAGGGTTTTAGCG

TTAGGGTTTTTAGCG

TTAGGGATAGCG

TTAGGGTAAGCG

TTAGGGTTATCG

68a

68b

68c

68d

68e

68f

68g

67a

67b

67c

67d

CGCTAA5' 3'

CCCTAA5' 3'

CCCTAA5' 3'

Ar =

Scheme 29. Ligation of oligodeoxynucleotides by photocycloaddition of anthracenes

zylalcohol and photorepair of thymine dimers. Contrary to the previoussections, there is no shielding effect or pre-organisation of reaction compo-nents. Selectivity and efficiency were achieved by conducting the desiredphotoreaction only in the vicinity of the sensitiser and thus by conversionof the diffusion-controlled process to an intramolecular reaction.58

The first such system reported (compound 69a) used a flavin group asthe photomediator and Lews-acidic zinc(ii)–cyclene as the binding site.59–61

The zinc(ii)–cyclene bound 4-methoxybenzyl alcohol 70, and upon irradia-tion, the excited flavin chromophore oxidised the alcohol to the correspond-ing aldehyde (Scheme 30).62 After the reaction, the flavin skeleton was inits reduced state and required re-oxidation before the next turn of the cycle.Aerial oxygen rapidly re-oxidised the flavin unit and the reaction went forth.Thus, the net reaction was ocidation of 4-methoxybenzylalcohol 70 to thecorresponding aldehyde by oxygen; however, in the absence of the template69a or light, the reaction did not proceed. The importance of the proximityof the substrate and the sensitiser was proved by experiments with flavinwhich did not contain the zinc(ii)–cyclene binding site or with equimolarmixtures of flavin and zinc(ii)-cyclene perchlorate. Only the complete sen-sitiser 69a which contained covalently bound zinc(ii)-cyclene exhibited high


O

HO

N

N

N

N

O

OO

N N

NNH

HH

Zn2+

R13

1 2

69 70.

OO

hν

2 e-2 e-

O2

H2O2

R1 = 69a

69b

70

Scheme 30. Schematic representation of the catalytic oxidation of 4-methoxybezyl alcoholby a flavin unit; Key: (1) irradiation, (2) oxidation of substrate,(3) reoxidation of flavin

conversion and turnover number (yield 90% after 2 hrs of irradiation). Thesensitiser acted as a true catalyst: 10 mol % of the sensitiser was suffi-cient for efficient oxidation of the substrate. Oxidation was possible bothin acetonitrile and aqueous buffer solution due to the good solubility of thesensitiser 69a and high association constant.

The sensitisers 69 were also found to act as an artificial functionalmodel of a photolyase.63 A major environmental damage to DNA is theUV-radiation-induced [2+2] cycloaddition between two adjacent thymineresidues on one strand of DNA, leading to formation of a cis-syn thyminecyclobutane dimer. DNA lyase selectively recognises the thymine dimers andrepairs them by photinduced electron transfer using a non-covalently boundreduced flavin as the electron donor, excited by visible light. The describedphotolyase model 69 selectively bound to the thymine dimers 71 both inorganic solvents (compound 69b with thymine dimer 71b) and aqueousenvironment (compound 69a with thymine dimer 71a), and upon irradia-tion by visible light induced electron-transfer-catalysed cycloreversion, thuscleaving the cyclobutane ring (Scheme 31). Analysis of the reaction mix-ture showed a fast and clean conversion of the thymine dimer 71b to themonomers 72b. The reaction was carried out in acetonitrile (completiontime 120 min) or methanol (completion time 60 min). High hydrophilicityof the glycol-substituted flavin sensitiser 69a allowed for efficient repair ofthe thymine dimer 71a in water. The reaction in water was faster than thatin organic solvents (78% conversion after 10 s of irradiation). This observa-tion was explained by higher polarity of water and thus by better stackingof the flavin unit and the thymine dimer facilitating the electron-transfer


N

O

O

N

N N

N

Zn2+

H

HH

ON

N

NH

N O

O

N N

NH

O

O

COOR2 COOR2

R

-

N

NH

COOR2

O

O

NH

N

O

O

COOR2

43

1

2-

OO

N

N

O

O

COOR2

[Zn2+] -

-

2

71

72a72b

R2 = HCH2Ph

71a71b

up to 100%up to 80%

R1 = 69a

69b

hν

Scheme 31. Functional model of photolyase; Key: (1) irradiation, (2) electron transfer,(3) cycloreversion, (4) back electron transfer

step.Flavin-based artificial photolyase models were studied in great detail.

Most of the examples featured flavin units covalently linked to the pyridiminedimers, such as in compound 73 (Scheme 32).64–66

The efficiency of the cycloreversion was, apart from a covalent link-age, also enhanced by incorporation of flavin into an oligopeptide strand.67

The synthesis made use of an artificial flavin-containing amino acid 74(Scheme 33), leading to flavin-containing oligopeptides 75 and 76. Tomimic a helix–loop–helix protein and enhance binding to DNA, two flavin-containing peptides were linked by a bis(bromoacetyl)benzene template toyield compound 76 (Scheme 34). In solution, the oligopeptides 75 and 76bound to a lesion-containing synthetic oligonucleotide 77, presumably by


NH

N N

NH

O O

OO

H H

H HO O

O ON

N

N

N

OO

N

N

NN

O

O

73

Scheme 32. Example of a photolyase model with covalently attached flavin units64f

N

NNH

N O

O

COOHNH2

Fl=

74

Scheme 33. Artificial flavin-containing amino acid and its schematic representation

interaction of the phosphodiester and arginine or lysine residues. Upon ir-radiation by daylight or 366 nm light, the flavin unit was reduced in thepresence of ethylene diamine tetraacetic acid and the excited reduced chro-mophore then cleanly repaired the DNA damage (Scheme 35).

Thymine dimer could be efficiently cleaved by a carbazole nucleoside in-corporated into a complementary DNA strand.68 Upon irradiation, cyclore-version of the thymine dimer occured, and the repaired oligodeoxynucleotidewas isolated with 92% yield. Templates with an additional carbazole nucle-oside or adenosine residue in the active site worked with comparable yields(94% and 93%, respectively). When mismatched bases were present in thedamaged strand, no repair was observed.

Thymine dimer was also selectively recognised by a compound contain-ing two 2,6-bis(acylamino)pyridine binding sites and an anthraquinone chro-mophore.69 It was also possible to prevent formation of thymine dimer by


O

O

S

S

NH2 O

FlNH2

O

Ac

Fl Ac

Fl Ac

AENVKS RRRERETAAKRRDA

CGGAENVKS RRRERETAAKRRDA

CGGAENVKS RRRERETAAKRRDA

75

76

Scheme 34. Flavin-containing oligopeptides

5'-CGCGT-U=U-TGCGC-3'

5'-CGCGT-U-U-TGCGC-3'

hν oligopeptide 75,76

77

up to 100%

Scheme 35. Photorepair of DNA damage by a flavin-containing oligopeptide

binding the two thymine bases to a dinuclear complex so far apart fromeach other that they could not undergo the photocycloaddition reaction.70

However, this is an example of efficient reaction blocking, not enhancement,and the work is therefore not discussed in detail.

6. Photochemical Reactions in a Molecular Flask

A variety of photochemical transformations are influenced and enhancedby the presence of a molecular flask, i.e., a large template molecule whichincludes the guest or guests, thus pre-organising them advantageously fora desired photochemical reaction. These premises are similar to those out-lined in Section 3, where photochemistry within non-covalent assemblies wasdescribed. In this section we discuss examples where the host, or templatemolecule, is significantly larger than the guest and can be therefore describedas a true molecular flask which protects substrates from the surrounding en-vironment and thus controls the course of a photochemical reaction.

6. Photochemical Reactions in a Molecular Flask 31

The spectrum of reactions enhanced in this manner is wide and rangesfrom dimerisations of aromatic and olefinic systems to selective photochem-ical cleavage or transport of a host-bound sensitiser toward a light beam.The molecular flasks used are known from other supramolecular applica-tions. Cyclodextrins are favourite template scaffolds because of their goodavailability, various sizes, and inherent chirality, a property which was usedfor enantioselective photochemical rections.71–74 Other examples includeself-assembled cage and bowl, cucurbiturils, and a self-assembled cavitand.In this section, we describe photodimerisations and photocycloaddition re-actions first, and continue with selective photochemical cleavage reactionsand systems with transport function.

The best studied photochemical reaction in a molecular flask is prob-ably the dimerisation of anthracene carboxylate 78.75–83 γ-Cyclodextrin(79) and modified (80–91) cyclodextrins were used to manipulate yieldand stereochemical outcome of the dimerisation reaction (Scheme 36). Twoof the possible four isomers 92a–d are chiral (Scheme 37), which stimu-lated investigations on the asymmetric induction by the inherently chiralcyclodextrin.43 Different experimental conditions disallow for direct com-parison of the individual results, but effects of the host geometry on theselectivity of the dimerisation reaction are definitely observed. Overall, theresults from photodimerisation of anthracene carboxylic acid in the presenceof γ-cyclodextrin-based templates can be summarised in the following way:The presence of any γ-cyclodextrin derivative accelerates the reaction by anincrease of the local concentration of anthracene molecules. The structureof the cycloadduct necessarily reflects the structure of the precursor ground-state assembly because within the excited singlet state lifetime (ca. 10 ns)the orientation of the aggregate does not change by dissociation and no re-action of free anthracene-2-carboxylic acid with the 1:1 inclusion complexoccurs. Therefore, better control of the relative geometry leads to higherselectivity of the photodimerisation. With γ-cyclodextrin 79, irradiation ofanthracene-2-carboxylate 78 led preferentially to formation of isomers 92a(up to 43% yield) and 92b (up to 46% yield and 41% ee).77 With bis-pyridinio-γ-cyclodextrins 80AB–AE, the electrostatic repulsion of the twocarboxylate groups was partly overcome and the two molecules aligned tothe pyridinium units.78 Preference for isomers 92a (up to 43% yield) and92c (up to 32% yield and ca. 10% ee) was observed. Diamino-modified γ-cyclodextrins 81AB-AE led to an increased formation of isomers 92c (upto 32% yield, 27% ee) and 92d (up to 21% yield) in comparison to the iso-mer distribution in the presence of γ-cyclodextrin 79.79 γ-Cyclodextrin 85,bearing a dicationic substituent, was selective for the formation of isomers92c (up to 42% yield, 41% ee) and 92d (up to 41% yield).80 γ-Cyclodextrinwith flexible hydrophobic caps 83 and 84 did not significantly influence theproduct distribution or enantioselectivity of the dimerisation when comparedto γ-cyclodextrin 79.81 However, γ-cyclodextrin 82 with a rigid hydropho-


N+

N+

NH3+

H3N+ NH NH

++ +

R2R1 SO3NaSOO

O

SOO

O

SSOO

OO

O

O

OO

OHOH

OH

7

O

OH

R1

O

R2

OO

OHOH

OH

7

O

OH

O

O

OO

OHOH

OH

O

R

OH

O

OH7

AB AC AD AE

OO

OHOH

OH

8

R1 = OHNH2

R2 = OTsOH

80 81 82

83

84

85

A D

88

R1 = R2 = OH

R1 = R2 =

R1 = = R2

γ-CD γ-CD γ-CD

79

=γ-CD

8687

R = N3NH2OH

899091

Scheme 36. γ-Cyclodextrins used as hosts for the photodimerisation of anthracene car-boxylate 78


COOH

COOH

HOOC

COOH

HOOC

COOH

COOH

COOH

COOH

2hν

92a 92b

92c 92d

chiral

chiral

78

hostsolvent

Scheme 37. Yields and ee values depend on host and conditions employed and arediscussed in text

bic cap enabled formation of product 92b with up to 58% ee, although theyield was lower than with γ-cyclodextrin 79. A series of secondary face84

modified γ-cyclodextrins 86–91 did not influence the relative yields of cy-clodimers 92a–d either.82 However, the enantioselectivity of the reactionwas significantly enhanced with some derivatives. Isomer 92b was obtainedwith 47% yield and 53% ee at 0 ◦C and 0.1 MPa and with 53% yield and71% ee at −21.5 ◦C and 210 MPa when using the modified γ-cyclodextrin90 as template.

Other investigations have concerned photodimerisation of anthracene-2-carboxylate 78 in the binding pocket of the protein bovine serum albumin,which leads to selective formation of isomers 92c (up to 38% yield, 41% ee)and 92d (up to 43% yield).83

A self-asembled cage-like molecule 93 (Scheme 38) was used for highlyselective photochemical dimerisation of acenaphthylenes 94 and 95,85,86

naphthoquinone (96)86 and antracene-9-carbaldehyde (97),85 which adopteda well defined geometry in the stringent environment. The self-assembledcage 93 was composed out of six metal complexes 98 and four tridentateligands 99a.

Irradiation of the acenapthylene molecules 94 and 95 (Scheme 39) acco-


N

N

N

N

N N

=

NH2

NH2

Pd[Pd] =

. 12 NO3-

N

NN

NN

N

[Pd]

[Pd][Pd]

[Pd]

[Pd]

[Pd]

N

N N

N

N

N

NN N

N

N

N

N

N N

N

N

N

98 99a

93

12+

Scheme 38. Self-assembled molecular cage

modated inside the cage 93 led to selective formation of the syn dimer 100and 101 in more than 98% yield and without any other regio- or stereoiso-mers.85,86

Similarly, photodimerisation of naphthoquinone 96a inside a self-assem-bled bowl-like molecule 102 consisting of six units of 98 and four units of99b led selectively to the formation of the syn dimer 103a (Scheme 40).86

The regioselectivity of the photodimerisation of 2-methylnaphthoquinone96b was very high (96% head-to-tail dimer 103b) with molecular cage 93but only moderate (78% head-to-tail dimer 103b) with molecular bowl 102.Dimerisation of 5-methoxynaphthoquinone 96c led to formation of isomer103c in 79% yield in the presence of molecular bowl 102. Irradiation of 96bwithout the cage (50 mm, 3 hrs, benzene) did not afford any dimerisationproduct, while that of 96c gave the anti dimer in 21% yield.

Irradiation of the complex of anthracene-9-carbaldehyde (97) and cage


R

R

R

100101

R = HMe

9495

100%, > 98%> 98%

hνD2O

93

Scheme 39. Photodimerisation of acenaphthylenes in the molecular cage 93

O

O

R1R2 O

O

O

O

R1

R1

R2

R2

N

N

N

N

N

NN

N

NN

N

N

NN

NN

N

N

[Pd]

NNN

N

N

NN

N N

NN

N

[Pd]

[Pd]

[Pd]

[Pd]

[Pd]. 12 NO3-

102

103a103b103c

Lb =

99b

> 98%78% (96% within cage 93)79%

102

hνD2O

96a96b96c

R1 = HMeH

HHOMe

R2 =

=

12+

Scheme 40. Photodimerisation of naphthoquinones in a molecular bowl 102


93 in a ratio 2:1 led to selective formation of the photodimerisation product104a (Scheme 41).85 The cage 93 accomodated the two guest in a geometry

CHO OHC

OHC

OHC

OHC

+

104a 104b97

93

100% not observed

hνD2O

Scheme 41. Photodimerisation of anthracene-9-carbaldehyde in molecular cage

leading to exclusive formation of the isomer shown; formation of the otherpossible isomer 104b was not observed. Other 9-substituted anthraceneswere photoinactive inside the cage.

Acenaphthylene 94 undergoes efficient cycloaddition with 5-ethoxynaph-thoquinone 105 using cage 93 (Scheme 42, top).87–89 Irradiation of the1:1:1 complex (acenaphthylene 94:naphthoquinone 105:cage 93) led to se-lective formation of cis-syn cycloadduct 106a in 92% yield. Formation ofthe ternary complex and hence the resulting photoproduct was governed bythe bulkiness of the naphthoquinone guest: when 5-methoxynaphthoquinone96c was used, the same isomer 106b was still the major product (yield 44%),but the reaction mixture contained 100 (6%) and 103c (22%) as side prod-ucts. With naphthoquinone (96a), selectivity was impaired even further(100, 21%; 106c, 35%; 103a, 14%). The cage was also successfully usedfor cycloaddition reactions of otherwise photoinactive guests (Scheme 42,bottom). N -Benzylmaleimide (107), a photochemically inert substrate, ef-ficiently undergoes cycloaddition with acenaphthylene (94) when inside themolecular cage 93 and yields the syn cycloadduct 108 exclusively in 97%yield. Similar efficiency was observed for reaction of N -arylmaleimides 109aand 109b with dibenzosuberenone 110. The syn cycloadducts 111a and111b were both isolated in quantitative yield.

The cage 93 and related molecular capsules were successfully used forstabilisation of photochemically generated reactive intermediates which thenunderwent desired reactions.90 However, in these cases, irradiation by lightwas only used to generate the intermediate, while the reactions of interestwere thermal, and these reactions are therefore not discussed in detail.

A reversal of the syn:anti selectivity of the Paternó–Büchi reaction ofadamantan-2-one 112 with fumaronitrile 113 was achieved by use of β-cyclodextrin.91 A series of adamantan-2-ones 112 reacted with fumaroni-trile 113 to yields the corresponding syn and anti oxetanes 114 and 115(Scheme 43) with some preference for the syn isomer. The syn:anti ratiolargely depends on the 5-substituent of the adamantanone 112. During


O

O

R

O

O

R O

O

O

O

R1

R1

HOMe

R1 =

NO O

Ph

N

O

O

Ph

O

ON

O

O

ArNO O

RR

+

R = OEtOMeH

10596c96a

100 106a-c

94

93

103a103c

+

R = MeOMe

109a109b

107 9493

108

110

93

111a111b

hνD2O

hνD2O

97%

hν

quantitativequantitative

Scheme 42. Cross-photocycloadditions of substrates in the molecular cage 93; See textfor the yields of acenaphthylene (94)–naphthoquinone (96a,c, 105) cycloadditions


CN

NC

O

R

2

5

R

OCN

CNR

O

CN

CN

O

CNNC

NC

CN

+ +

antisyn

hνH2Oβ-CD

113 114 115R = FClBrOHPht-Bu

112a112b112c112d112e112f

Scheme 43. Paternó–Büchi reaction of adamantan-2-ones with fumaronitrile in the pres-ence of β-cyclodextrin (β-CD) favouring formation of the syn isomer 114

the reaction the substrate is firmly anchored in the β-cyclodextrin cavity,thus blocking approach of fumaronitrile from one face of the carbonyl group.Better fitting substituents such as phenyl (112e) or tert-butyl (112f) leadpreferentially to the syn cycloadduct 114 with ratios up to 86:14. In theabsence of β-cyclodextrin, selectivity for the anti isomer 115 (e.g. R = Cl,syn:anti, 43:57; R = t-Bu, syn:anti, 36:82) is observed in aqueous solution.The presence of α- or γ-cyclodextrin does not change this preference signif-icantly.92

6-(5-Cyanonaphthyl-1-carboamido)-6-deoxy-β-cyclodextrin was found tomanipulate the anti-Markovnikov addition of methanol to 1,1-diphenylprop-ene.93 In 50% aqueous methanol, 2-methoxy-1,1-diphenylpropane was ob-tained with up to 61% yield and 6% ee (1 eq. of the template, −40 ◦C).In pure methanol, conversion below 1% was observed. In 25% aqueousmethanol, the isolated yields were 15–18% with ee values of up to 11%.

Photocycloaddition of double bonds in stilbenes94–96 and its nitrogen-containing analogues,97 2-aminopyridine,98 and cinnamates99,100 was stud-ied in the presence of cyclodextrins or cucurbiturils 116 (Scheme 44). Whencucurbituril 116 of the correct size was used, the guest molecules coordinatedto the portal oxygens atoms in a defined geometry, leading to high selectivityfor some of the products. In the case of diaminostilbenes 117, photodimeri-sation in the cavity of cucurbit[8]uril 116c occured and syn cycloadduct124 was selectively formed with a syn:anti ratio higher than 95:5.95 In theabsence of the template 116c, the main reaction pathway was the isomeri-sation to the corresponding Z -isomer 137. In case of dimethylaminomethyl-


Ar2Ar1

NH3+

NH+

NH+

NH+

NH+

NH+

Ar2Ar1 Ar2

Ar1

OH

Ar1

Ar2

Ar2

Ar1

Ar1

Ar2

Ar2

Ar1

Ar1

Ar2

CH2

N

NN

N N

O

O

H HCH2

CH2

n

Ar1 = Ar2 =

Ar1 = Ar2 =

Ar1 = Ar2 =

Ar1 = Ar2 =

Ar1 = Ph Ar2 =

Ar1 = Ar2 =

117 124

118 125

119

120

121

122

123

137 138 139

126

127

128

129

130

131

132

133

134

135

136

n = 678

116a116b116c

Scheme 44. Photodimerisation of stilbene and its nitrogen-containing analogues in thecavity of cucurbit[8]uril 116c or cyclodextrins


substituted stilbene 118, irradiation in the presence of γ-cyclodextrin ledto selective formation of the trans dimer 131 (yield 79%, minor cis dimer125 19% yield).101 In the presence of the smaller α- or β-cyclodextrins or inthe absence of any template, the photodimerisation reaction was prohibited.Similar enhancement was observed for the nitrogen-containing stilbene ana-logues 119 to 123.97 Irradiation in the absence of any template yielded thecorresponding cis isomers 137, an intramolecular cyclisation product 138,and the hydration product 139, while in the presence of cucurbit[8]uril 116ca mixture of the corresponding syn dimers 126–130 (81–90% yield) and antidimers 132–136 (0–6% yield) was obtained. The cavity of cucurbit[7]uril116b was too small to accommodate two molecules of the guest, and itspresence did not enhance dimerisation selectivity.

The [4+4] photocycloaddition of 2-aminopyridine hydrochloride 140 in-side the cavity of cucurbit[7]uril 116b gives an increased selectivity for theanti-trans dimer 141a (yield up to 90% without any side products).98 In theabsence of the template, both anti-trans 141a and syn-trans 141b dimerswere formed in a ratio of ca. 4:1 (Scheme 45). The template also protectedthe protonated form from thermal dissociation to the starting materials.

NH

+NH2

NH

+

NH

+

NH2NH2

NH

+

NH

+

NH2NH2

2

anti-trans syn-trans

hν

CB7 116b

140141a 141b

90% not observed

Scheme 45. Photodimerisation of 2-aminopyridine hydrochloride in the cavity of cucur-bit[7]uril 116b

Irradiation of trans-cinnamic acid 142 included in the cavity of cucur-bit[8]uril 116c led to selective formation of the cyclic dimer 143 togetherwith the corresponding cis isomer 144 (Scheme 46).99 Best results were ob-served for 4-amino-trans-cinnamic acid hydrochloride 142c, which yielded88% of the corresponding dimer 143c.

Enantiodifferentiating photoisomerisation of Z -cyclooctene 145 to chi-ral E -cyclooctene 147 (Scheme 47) was achieved in the cavity of 6-O -substituted cyclodextrins 146a–g.102–104 Host compound 146b gave thebest results with E/Z ratio of 0.29 and ee of 24% after irradiation at −40 ◦Cin 50% aqueous methanol. With the other cyclodextrin derviatives 146a,c–g, higher E/Z ratios were observed but the ee values remained low. Pho-toisomerisation of Z -cyclooctene 145 in the presence of a permethylated6-O -benzoyl-β-cyclodextrin was recently described.105 This host is more


R1

COOH

R2

Ar

Ar

COOH

COOHAr COOH+

142 143 144

hν

CB8 116c

abcdef

R1 = R2 = HR1 = OH

NH3+Cl-OMeHH

R2 = HHHOMeMe

54%38%88%72%72%83%

Scheme 46. Photodimerisation of cinnamic acid in the cavity of cucurbit[8]uril 116c

OO

OHOH

OH

n

HN

NH

O

O O

R

=

145 147

+hν6-O-modified cyclodextrins146 as chiral sensitising hosts

146gn = 677778

R = HHo-COOMem-COOMep-COOMeH

146a146b146c146d146e146f

n = 7

Scheme 47. Enantiodifferentiating photoisomerisation of Z -cyclooctene; E/Z ratios andee values depend on the chiral host 146 and conditions (see text)


flexible due to the absence of the hydrogen-bond network on the secondaryface84 of the cyclodextrin. The E/Z ratios and ee values critically depend onthe temperature and composition of the solvent. In pure methanol, the E/Zratio remained low with no significant asymmetric induction since practi-cally no complexation of the Z -cyclooctene 145 by the host occurred andall photochemistry took place “outside” the cavity, in the bulk media. Inaqueous solutions containing 50% or 25% of methanol, the chiral sense ofthe photoproduct 147 could be switched by temperature. The best resultsreached 9% ee for for (R)-(−)-147 (obtained in 50% aqueous methanol at−40 ◦C) and 4% ee for (S)-(+)-147 (obtained in 25% aqueous methanol at40 ◦C).

The selectivity of photochemical cleavage reactions was successfully ma-nipulated by inclusion of the starting material in a self-assembled cavitand106

or in a water-soluble calixarene.107 In the absence of a template, photo-chemical cleavage of 1-phenyl-3-p-tolylacetone 148a yielded a mixture ofdecarbonylation products 149a–c (Scheme 48).106 When included in a self-assembled cavitand 1522, ketone 148a gave a mixture of the decarbonylatedproduct 149a (41%), rearranged decarbonylated products 150a,c,e (15%),and rearranged product 151a (44%). 1,3-Diphenylacetone 148b renderedsimilar results (decarbonylated product 149b in 38% yield, rearranged de-carbonylated products 150b,d in 13% yield, and rearranged product 151bin 49% yield) unlike 1,3-bis-(p-tolyl)-acetone 148c whose geometry in thecavity favoured formation of the decaronylated product 149c in 96% yieldalong with minute amounts of side products.

Benzoin alkyl ethers 153 (Scheme 49) preferentially undergo NorrishType II cleavage when encapsulated in the cavity of p-sulphonatocalixarenes154.107 The substrate was locked by the cavity in a conformation whichfavoured the γ-hydrogen abstraction (yielding deoxybenzoin as the majorproduct) rather than the cleavage of the C–C bond (yielding pinacol etheras the major product). Higher yields of the Norrish Type II reaction wereobserved with the larger calixarene 154b: benzoines 153a–b yielded upto 96% of deoxybenzoin 155 and benzoin 153c up to 85% of deoxybenzoin155 together with minor amounts of the corresponding pinacol ethers 156a–c. The smaller cavity of calixarene 154a allowed cage escape, and observedyields of the the Type II product were therefore lower. Compound 153a gaveup to 70% of deoxybenzoin 155 and 30% of pinacol ether 156a. Compound153b yielded up to 67% of 155 and 32% of 156b, and 153c gave up to65% of 155 and 35% of 156c.

Compound 158 is able to transport sensitiser 157 and release the phtha-locyanine upon irradiation.108,109 Two cyclodextrin units are linked in 158by a tether containing a central C=C double bond (Scheme 50). The tert-butyl groups of the phthalocyanine sensitiser 157 bind into the cyclodextrincavities, making complex 157·158 water-soluble. Upon irradiation, sensi-tiser 157 effected formation of singlet oxygen110,111 and thus cleavage of


OR1 R2

R1

R2

R1

R1

R1R2

R1O

R1 = CH3, R2 = HR1 = R2 = HR1 = R2 = CH3

148a148b148c

R1 = CH3, R2 = HR1 = R2 = HR1 = R2 = CH3

149a149b149c

R1 = CH3H

150a156b

R1 = CH3H

150c150d

R1 = CH3, R2 = HR1 = R2 = CH3

150e150f

R1 = CH3H

151a151b

h ν,152 or no template

152

Scheme 48. Photochemical cleavages of substituted acetones; See text for yields of indi-vidual products. Structure of compound 152 reproduced from ref. 106


n

SO3Na

OH

Ph

O

Ph

OR

O

PhPh

PhORH

PhHRO

hν +

155156a-cR =

n = 68

154a154b

MeEti-Pr

153a153b153c

Scheme 49. Selective Norrish type II cleavage in the cavity of a calixarene; See text foryields of individual products

the C=C double bond yielding 159. Once the two cyclodextrin units weredisconnected, the binding affinity for the sensitiser 157 decreased. The co-operative effect of the two cyclodextrins is lost, and the loose linker chainscompete for binding into the cyclodextrin cavity. The complex thereforedissociates, leading to precipitation of the now insoluble phthalocyanine sen-sitiser. Eventually, the sensitiser concentrated in a directed light beam, asshown by an experiment with a reaction irradiated through a small hole ina shield surrounding the reaction vessel. Binding to the cyclodextrin dimerand rate of the cleavage reaction was improved in a series of zinc phthalo-cyanines.112

7. Conclusion

The concept of reaction control in homogeneous solutions by templates isclearly established in photochemistry. Shielding of prochiral faces, topo-chemical reaction control, templation, and aggregate or inclusion complexformation are strategies found to enhance and control reaction efficiency,regio- or stereoselectivity. However, the number of truly catalytic examplesand their efficiency remain limited. Development of catalytic templatescontrolling the stereochemistry of photoreactions with high precision andactivity will be therefore a future challenge in the field.

8. Notes and References

[1] Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191–1223.[2] For examples, see: (a) Natarajan, A.; Mague, J. T.; Ramamurthy,

V. Crystal Growth & Design 2005, 5, 2348–2355. (b) Natarajan, A.;Mague, J. T.; Ramamurthy, V. J. Am. Chem. Soc. 2005, 127, 3568–3576. (c) Alam, M. M. Synlett 2003, 1755–1756.

8. Notes and References 45

SNH

O

S SH H

NH

OS

NH

N

NH

N

N N

NN

O

O

O

O

SO

O

OH

SNH

O

S

O

SNH

OS

O

3

1

23O2

1O2

hν

1O2

157

158

β

β

β

β

159

Scheme 50. Cyclodextrin-based photosensitiser carrier 158; Yield of the reaction wasnot given. Key: (1) irradiation, (2) sensitisation of oxygen, (3) cleavage of C=C doublebond


[3] For an example, see: Vicario, J. L.; Badia, D.; Carrillo, L.; Reyes, E.;Etxebarria, J. Curr. Org. Chem. 2005, 9, 219–235.

[4] For an example, see: Noyori, R. Angew. Chem., Int. Ed. 2002, 41,2008–2022.

[5] Schmidt, G. M. J. Pure Appl. Chem. 1971, 27, 647–678.[6] In the classical example, photodimerisation of cinnamic acid in the

crystal, a limiting distance of 0.42 nm was defined. If groups are closer,they react. If groups are further apart, no reaction is observed.

[7] For recent reviews on solid-state control of photoreactivity, see: (a)Ito, Y. Synthesis 1998, 1–32. (b) Feldman, K. S.; Campbell, R. F.;Saunders, J. C.; Ahn, C.; Masters, K. M. J. Org. Chem. 1997, 62,8814–8820.

[8] One example is the photodimerisation of crystalline anthracene, whichoccurs at a “topochemically forbidden” distance of the reaction groupsof 0.6 nm. See: Kaupp, G. Angew. Chem., Int. Ed. 1992, 31, 595–598and references therein.

[9] (a) Ramamurthy, V.; Schanze, K. S. Photochemistry of OrganicMolecules in Isotropic and Anisotropic Media; Marcel Dekker, Inc.:New York, 2003. (b) Kaupp, G. In Comprehensive SupramolecularChemistry ; Davies, J. E., Ripmeester, J. A., Eds.; Elsevier: Oxford,UK, 1996; Vol. 6, p 381. (c) Kaupp, G. Adv. Photochem. 1995, 19,119.

[10] For a review on the chemistry of host–guest inclusion complexes, see:Toda, F. In Comprehensive Supramolecular Chemistry ; Macnicol, D.D., Toda, F., Bishop, R., Eds.; Elsevier: Oxford, UK, 1996; Vol. 6, p465.

[11] For example, see: Toda, F.; Miyamoto, H. J. Chem. Soc., PerkinTrans. 1993, 1, 1129–1132.

[12] For reviews on zeolites as supramolecuar hosts for photochemicaltransformations, see: (a) Ramamurthy, V.; Natarajan, A.; Kaanu-malle, L. S.; Karthikeyan, S.; Sivaguru, J.; Shailaja, J.; Joy, A. Mol.Supramol. Photochem. 2004, 11, 563–631. (b) Sivaguru, J.; Natarajan,A.; Kaanumalle, L. S.; Shailaja, J.; Uppili, S.; Joy, A.; Ramamurthy,V. Acc. Chem. Res. 2003, 36, 509–521. (c) Ramamurthy, V.; Garcia-Garibay, M. A. In Comprehensive Supramolecular Chemistry ; Alberti,G., Bein, T., Eds.; Elsevier: Oxford, UK, 1996; Vol. 7, p 693.

[13] For reviews on related topics exceeding the scope of this review, see:(a) Vriezema, D. M.; Aragones, M. C.; Elemans, J. A. A. W.; Cor-nelissen, J. J. L. M.; Rowan, A. E.; Nolte, R. J. M. Chem. Rev. 2005,105, 1445–1489. (b) Wada, T.; Inoue, Y. Mol. Supramol. Photochem.2004, 11, 341–384. (c) Inoue, Y. Mol. Supramol. Photochem. 2004,11, 129–177. (d) Cooke, G. Angew. Chem., Int. Ed. 2003, 42, 4860–4870. (e) Tung, C.-H.; Wu, L.-Z.; Zhang, L.-P.; Chen, B. Acc. Chem.Res. 2003, 36, 39–47. (f) Armaroli, N. Photochem. Photobiol. Sci.

8. Notes and References 47

2003, 2, 73–87. (g) Shipway, A. N.; Willner, I. Acc. Chem. Res. 2001,34, 421–432. (h) Ito, Y. Synthesis 1998, 1–32.

[14] For recent reviews on topics discussed in this review, see: (a) Wessig,P. Angew. Chem., Int. Ed. 2006, 45, 2168–2171. (b) Huang, C.-H.;Bassani, D. M. Eur. J. Org. Chem. 2005, 4041–4050.

[15] For examples on the influence of hydrogen bonding on intramolecularof non-templated photoreactions, see: (a) Sieburth, S. M.; McGee,K. F., Jr.; Al-Tel, T. H. J. Am. Chem. Soc. 1998, 120, 587–588. (b)Crimmins, M. T.; Choy, A. L. J. Am. Chem. Soc. 1997, 119, 10237–10238. (c) Sieburth, S. M.; Joshi, P. V. J. Org. Chem. 1993, 58,1661–1663.

[16] For a review on non-templated asymmetric photochemical reactionsin solution, see: Inoue, Y. Chem. Rev. 1992, 92, 741–770.

[17] Mori, K.; Murai, O.; Hashimoto, S.; Nakamura, Y. Tetrahedron Lett.1996, 37, 8523–8526.

[18] The example is in principle similar to the following: (a) Amirsakis, D.G.; Elizarov, A. M.; Garcia-Garibay, M. A.; Glink, P. T.; Stoddart,J. F.; White, A. J. P.; Williams, D. J. Angew. Chem., Int. Ed. 2003,42, 1126–1132. (b) Beak, P.; Zeigler, J. M. J. Org. Chem. 1981, 46,619–624.

[19] Bach, T.; Bergmann, H.; Harms, K. J. Am. Chem. Soc. 1999, 121,10650–10651.

[20] Bach, T.; Bergmann, H.; Brummerhop, H.; Lewis, W.; Harms, K.Chem. Eur. J. 2001, 7, 4512–4520.

[21] Bach, T.; Bergmann, H.; Harms, K. Angew. Chem., Int. Ed. 2000,39, 2302–2304.

[22] (a) Bach, T.; Bergmann, H.; Grosch, B.; Harms, H. J. Am. Chem. Soc.2002, 124, 7982–7990. (b) Bach, T.; Bergmann, H. J. Am. Chem. Soc.2000, 122, 11525–11526.

[23] Selig, P.; Bach, T. J. Org. Chem. 2006, 71, 5662–5673.[24] (a) Grosch, B.; Orlebar, C. N.; Herdtweck, E.; Kaneda, M.; Wada, T.;

Inoue, Y.; Bach, T. Chem. Eur. J. 2004, 10, 2179–2189. (b) Grosch,B.; Orlebar, C. N.; Herdtweck, E.; Massa, W.; Bach, T. Angew. Chem.,Int. Ed. 2003, 42, 3693–3696. The reaction itself was thermal, notphotochemical, and is therefore not discussed in detail.

[25] Dressel, M.; Bach, T. Org. Lett. 2006, 8, 3145–3147.[26] Bach, T.; Bergmann, H.; Harms, K. Org. Lett. 2001, 3, 601–603.[27] (a) Bach, T.; Aechtner, T.; Neumüller, B. Chem. Eur. J. 2002, 8,

2464–2474. (b) Bach, T.; Aechtner, T.; Neumüller, B. Chem. Com-mun. 2001, 607–608.

[28] Bach, T.; Grosch, B.; Strassner, T.; Herdtweck, E. J. Org. Chem.2003, 68, 1107–1107.

[29] Bauer, A.; West

dissertation jiri svoboda - uni-regensburg.de · 2011. 7. 20. · *this chapter was already...

Documents